Previous research on the natural history of smoking has focused on overall group trajectories without considering the possibility of risk subgroup variation. To address this limitation, the authors of the present study aimed to identify subgroups with varying trajectories of smoking behavior. The authors accomplished this within a cohort-sequential study of a large community sample (N = 8,556) with measurements spanning ages 11-31. After removing 2 a priori groups (abstainers and erratics), the authors empirically identified 4 trajectory groups--early stable smokers, late stable smokers, experimenters, and quitters--and psychosocial variables from adolescence and young adulthood were significantly distinguished among them. Given recent advances in quantitative methods, it is now feasible to consider subgroups of trajectories within an overall longitudinal design.